Identification of potential antiviral compounds against SARS-CoV-2 structural and non structural protein targets: A pharmacoinformatics study of the CAS COVID-19 dataset. (June 2021)

• Record Type: Journal Article
• Authors: García, Rolando; Hussain, Anas; Koduru, Prasad; Atis, Murat; Wilson, Kathleen; Park, Jason Y.; Toby, Inimary; Diwa, Kimberly; Vu, Lavang; Ho, Samuel; Adnan, Fajar; Nguyen, Ashley; Cox, Andrew; Kirtek, Timothy; García, Patricia; Li, Yanhui; Jones, Heather; Shi, Guanglu; Green, Allen; Rosenbaum, David
• Is Part Of: Computers in biology and medicine. Volume 133(2021)
• Publication Date: 2021-06
• Subjects: SARS-CoV-2 -- COVID-19 -- Molecular docking -- Molecular dynamics; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2021.104364
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: SARS-CoV-2 is a newly discovered virus which causes COVID-19 (coronavirus disease of 2019), initially documented as a human pathogen in 2019 in the city of Wuhan China, has now quickly spread across the globe with an urgency to develop effective treatments for the virus and emerging variants. Therefore, to identify potential therapeutics, an antiviral catalogue of compounds from the CAS registry, a division of the American Chemical Society was evaluated using a pharmacoinformatics approach. A total of 49, 431 compounds were initially recovered. After a biological and chemical curation, only 23, 575 remained. A machine learning approach was then used to identify potential compounds as inhibitors of SARS-CoV-2 based on a training dataset of molecular descriptors and fingerprints of known reported compounds to have favorable interactions with SARS-CoV-2. This approach identified 178 compounds, however, a molecular docking analysis revealed only 39 compounds with strong binding to active sites. Downstream molecular analysis of four of these compounds revealed various non-covalent interactions along with simultaneous modulation between ligand and protein active site pockets. The pharmacological profiles of these compounds showed potential drug-likeness properties. Our work provides a list of candidate anti-viral compounds that may be used as a guide for further investigation and therapeutic development against SARS-CoV-2. Highlights: Machine learning methods, molecular docking and molecular dynamics identified candidate compounds to treat SARS-CoV-2. Compounds from the CAS COVID-19 dataset showed affinity to the spike protein and replication apparatus of SARS-CoV-2. Compounds showed combined therapeutic effect to structure, replication apparatus, while possibly inhibiting viral variants.